Ortiz de Montellano Paul R.(Ed.) Cytochrome P450. Structure, Mechanism, and Biochemistry

Подождите немного. Документ загружается.

Inhibition of Cytochrome P450 Enzymes

285

for these reactions but experimental evidence to

support the mechanisms is not available.

The P450 destruction mediated by halocarbons

was once believed to stem from the secondary

action of the lipid peroxides that are concomi-

tantly formed, but it is now evident that sub-

stances like CCI4 can destroy the heme group

directly^^^'

^^^' 423^26

J^IQ

associated cross-linking

of heme fragments to the protein suggests that a

radical species (CCI3 or CCI3O2) may be respon-

sible for heme destruction^^ ^ On the other hand,

the catalytic reduction of halocarbons, including

CCI4,

produces semistable complexes with

Soret maxima in the 450-500 nm range'^^^"^^^

Studies with model iron porphyrins, including the

detailed characterization of a dichlorocarbene-

metalloporphyrin complex"^^^, suggest that the

long-wavelength Soret bands are due to ferrous

halocarbene-heme complexes. This hypothesis is

supported by the finding that carbon monoxide is

formed in the reductive metabolism of CCI4 by

P450.

This reaction is likely to occur by a mecha-

nism similar to that for the generation of carbon

monoxide from methylenedioxyphenyl complexes

(Section 3.2.1)"^^^. In fact, porphyrin dichlorocar-

bene-iron complexes do react with water to give

carbon monoxide and with primary amines to give

isonitriles"^^^'

^^^.

Studies with halothane suggest

that it is also possible to form complexes in which

the halocarbon is a-bonded to ferric heme iron

atom435-^37

The links between formation of an iron-alkyl

complex and irreversible destruction of the heme

moiety have not been forged, but model studies

with diaryl- and carbethoxy-substituted carbene

complexes suggest that the halogenated carbenes

may shift to form a bond with a nitrogen of the

porphyrin'^^^^'^^ The resulting A/-haloalkyl adduct

are likely to undergo water-dependent hydrolysis

and might therefore not be detected by the methods

used to isolate other iV-alkyl porphyrins. However,

the formation of alternative reactive species that

attack the protein or the heme cannot be ruled out.

High (1-5 mM) concentrations of indo-

methacin and other nonsteroidal anti-inflamma-

tory agents reportedly denature P450 enzymes

because of their surfactant properties"^"^^. The

loss of P450 content seen when indomethacin is

added to liver microsomes is paralleled by essen-

tially stoichiometric appearance of a P420 peak.

Although it is likely that other agents cause P450

denaturation, it is unlikely that the process is phys-

iologically relevant because of the high drug

concentrations that are required.

4. P450 Enzyme Specificity

The isoform-specific inhibition of P450

enzymes is a promising avenue for the develop-

ment of therapeutic, insecticidal, and herbicidal

agents, as well as for investigation of the struc-

tures,

mechanisms, and biological roles of

individual P450 enzymes. The biosynthetic P450

enzymes have been the primary focus of efforts

to develop isoform-specific P450 inhibitors

because (a) they are better targets for specific

inhibitors because of their high substrate speci-

ficity and (b) there is high practical utility for such

inhibitors. In contrast, the broad, overlapping,

specificities of xenobiotic metabolizing P450

isoforms makes the design of isoform-specific

rather than -selective inhibitors more dififi-

cult'^'^^' ^^. Selective inhibitors of P450 enzymes,

as illustrated by the amphetamines^^^' ^^^,

TA0236,237,244-246^ Secobarbital^^' ^^\ gestodene^^^

ftirafylline^^^'

^^2,

1-ethynylpyrene^^' ^^, and

2,3',4,5'-tetramethoxystilbene'*'^^ are fairly com-

mon (see Appendix). Caution is required in

evaluating claims of inhibitor selectivity or speci-

ficity, as they are limited by the range of P450

enzymes actually examined. Only in the case where

an inhibitor has been tested with all the known

P450 isoforms in an organism can it be truly said to

be specific, at least in that organism. The claim for

specificity of inhibitors tested against only two or

three isoforms is necessarily limited.

5. Inhibitors of Biosynthetic

Enzymes

Several comprehensive review articles have

discussed the potential clinical relevance and

applications of inhibitors of biosynthetic P450

enzymes^' 446-448 jj^^ following discussion will

therefore be limited to an illustration of the

strategies employed in the design and develop-

ment of the currently available and/or prospective

inhibitors and their mechanistic diversity.

286

M.A.

Correia

and P.R.

Ortiz

Montellano

5.1.

P450,,,

A single P450 enzyme (P450^^^, CYPl 1 A) cat-

alyzes the three oxidative steps required to cleave

the side chain of cholesterol. A rational approach

to the development of P450g^^ inhibitors was

based on the incorporation of amino"^"^^^^^ and

thiol^^^ functions on the cholesterol side chain

at positions that favor their coordination to the

prosthetic heme iron (Figure 7.29), yielding

potent reversible inhibitors

(K^

= 25-700 nM).

Thus,

replacement of the first hydroxyl group

catalytically inserted into the cholesterol side

chain with an amine function yields (22R)-22-

aminocholesterol, one of the most potent P450g^^

inhibitors"^^^. The stereochemistry of this insertion

is critical for inhibition since (225)-22-aminocho-

lesterol, binds to P450g^^ —1,000 times more

weakly (^i = 13

|LJLM)

even though the amino

function is located on the correct carbon.

Various mechanism-based, irreversible inhi-

bitors of

P450g^^,

such as analogs of pregnenediol

with an acetylenic group grafted into their side

chain, have been developed (Figure 7.30)^^^^^^^.

Although this enzyme inactivation results in

destruction of the heme chromophore, no alkylated

heme was detected. Replacement of the side-chain

carbons beyond C-23 in 20-hydroxycholesterol by

a trimethylsilyl group also results in a P450g^^

mechanism-based inactivator (Figure 7.30)^^^.

Model studies have demonstrated that

1-substi-

tuted 3-trimethylsilyl-l-propanol is oxidized by

chemical reagents to ethylene, the trimethylsilyl

radical, and an aldehyde"^^^:

RCH(OH)CH2CH2Si(CH3)3 -> RCHO

+ CH2=CH2 + (CH3)3Si-

If the chemical model is relevant, P450g^^ may be

inactivated by reaction of the enzyme with the

trimethylsilyl radical or, less likely, from oxidation

of the ethylene produced in the initial catalytic

turnover. An interesting variant of a mechanism-

based inhibitor is provided by (20»S)-22-«or-22-

thiacholesterol, in which the sulfur that replaces

the carbon at position 22 is oxidized by P450g^^ to

a sulfoxide that is a potent but not an irreversible

inhibitor of the enzyme^^^' ^59,46o

Figure

7.29.

Reversible

inhibitors

biosynthetic

enzymes

are

usually

constructed

incorporating

nitrogen

other

coordinating

atom

into

hpophiHc

structure

with high

affinity

for the

protein

active

site

(see

Figure

7,1). An

inhibitor

of P450^^^ was

thus

obtained

placing

amino

group

at the

position

normally

occupied

by the

first

hydroxyl

group

added

the

cholesterol

side

chain

P450

5.2. Aromatase

Aromatase, through a three-step catalytic trans-

formation, controls the conversion of androgens to

estrogens. Competitive and mechanism-based

inhibitors of aromatase have been clinically

exploited in the treatment of estrogen-dependent

SiMe^

Figure

7.30. Two

mechanism-based

inhibitors

cytochrome

P450

Inhibition of Cytochrome P450 Enzymes

287

mammary tumors^' ^61^66

^j^^

benign prostatic

hyperplasia"^^^'

^^^,

and have some promise in the

control of coronary heart disease"^^^. Indeed, some

of the more promising newer agents are in cUnical

^j^^jg463^66 Aminoglutethimide, an inhibitor of

aro-

matase, has been used to treat hormone-dependent

metastatic breast carcinoma, but its poor specificity

and the incidence of side effects, primarily

from inhibition of P450g^^, has compromised its

utility^^^^^^. Replacement of the aminophenyl

group in aminoglutethimide (Figure 7.31) by a

p)nidine moiety affords [pyridoaminoglutethimide

(3-ethyl-3-(4-pyridyl) piperidine-2,6-dione)], an

agent that inhibits aromatase but not P450g^^'^^^'

'*^^. The enhanced specificity for aromatase may

reflect a differential positioning of the pyridyl

nitrogen within the two active sites that enables it

to coordinate with the heme of aromatase but not

of P450g^^. Interestingly, an aminoglutethimide

analog with a nitrogen at the opposite end is a

more potent P450g^^ inhibitor than glutethimide

but has little or no activity against aromatase^^^.

These findings suggest that the 19-methyl and

carbon 22 of the sterol side chain are separated by

a distance roughly equal to the length of the

aminoglutethimide structure. It is therefore

tempting to speculate that aminoglutethimide

binds within the active sites of P450 and

sec

P450^j.Q^ in approximately the same orientation as

the normal sterol substrate, and that the location

of the nitrogen in the two inhibitors dictates their

differential enzyme selectivity.

Potent inhibitors have also been developed that

use an imidazole or related functions to coordinate

to the aromatase iron atom. The most promising of

these for the treatment of breast cancer are fadro-

zole,

{4-(5,6,7,8-tetrahydroimidazo[l ,5a]-pyridin-

5-yl)benzonitrile monochloride} (CGS16949;

Figure 7.31), and its congener [bis-(p-

cyanophenyl)-imidazo-1 -yl-methane hemisucci-

nate]^^^^^^.

Both of these agents selectively inhibit

aromatase rather than of P450g^,^,

P45021,

P45011B475,478 Fadrozole does inhibit aldosterone

production (18-hydroxylase activity) in rats, but

this is much less pronounced with its congener^^^.

Phase I clinical trials of fadrozole indicated that it

is a potent inhibitor of estrogen biosynthesis in

postmenopausal women with advanced breast can-

^gj474-478 jj^ addition to being well tolerated by

patients, even at the maximally effective dose, it did

not significantly alter Cortisol, androstenedione,

testosterone or aldosterone levels"^^"^^^.

H2N

S O

X= CH2SCH3,N3, / \, / \

Aminoglutethimide

CH3N, .,N

R-76713

Fadrozole

:%:

N^y

Me Me

Anastrozole

Figure 7.31. Structures of some competitive inhibitors of aromatase. In each of these compounds, a heteroatom

is placed so that it can coordinate to the heme iron atom.

288

M.A. Correia and P.R. Ortiz de Montellano

However, fadrozole may now be surpassed by letro-

zole (CGS20267 or Femara; Figure 7.31), an

advanced nonsteroidal aromatase inhibitor, which

appears to be more potent and effective than fadro-

zole in the treatment of postmenopausal women

with advanced breast cancer^^^'

'*^^.

6 [(4-Chlorophenyl)( 1H-1,2,4-triazol-1 -yl)

methyl]-1 -methyl-1 H-benzotriazole, (R76713;

Figure 7.31), is a relatively selective and very

potent inhibitor of human placental aro-

matase"^^ ^~^^^. Its (+) enantiomer has a lower IC5Q

value than the (-) enantiomer when assayed

against human placental aromatase and exhibits

no appreciable inhibition of other steroidogenic

enzymes or liver microsomal P450 enzymes at

concentrations up to

1,000-times

the aromatase

Tp 482,483

Several other nonsteroidal compounds have

been developed as novel and selective aromatase

inhibitors, including 4-(4'-aminobenzyl)-2-oxazo-

lidinones"^^"^, 7-(alpha-azolylbenzyl)-1 H-indoles

and indolines of which l-ethyl-7-[(imidazol-l-yl)

(4-chlorophenyl)methyl]-lH-indole 12c exhibited

the most promising potency*^^, 4-imidazolylfla-

vans"^^^,

and anastrozole (Arimidex; Figure 7.31)"^^^.

Of these anastrozole has recently been approved

in the United States and several other countries for

the adjuvant treatment of postmenopausal women

with hormone receptor-positive early breast

cancer^^^'

^^^.

Powerful competitive steroid inhibitors of

aromatase have been synthesized by replacing the

C-19 methyl with sulfur- or nitrogen-containing

functions that can coordinate to the heme iron

(e.g.. Figure 7.31)^^^^^^ The 19-methyl substit-

uents include the following: CH^SCR^'^^^^^,

CH3SCH2CH/93^ HSCH2^9i, 494^ RSSCH2 (Org-

30958,

R = ethyl is best)^^^, SH^^i, NH2, and

NH2CH2^^^ Oxygen (oxiranyl), sulfur (thiiranyl),

and nitrogen (aziridinyl) three-membered rings

have also been used to replace the 19-methyl group

(Figure 7.32)^^^-^^^ the best of the resulting

steroids exhibiting

values in the

range.

The

epoxide, episulfide, and aziridine functions inhibit

the enzyme by stereoselective coordination of the

heteroatom to the iron atom but, despite their reactiv-

ity, apparently do not inactivate the enzyme.

However, the steroids in wiiich the 19-methyl group

is replaced by a sulfliydryl or HSCH2 group

(Figure 7.32) are irreversible mechanism-based

inhibitors rather than simple competitive

inhibitors'*^*.

The strategies developed for the inactivation of

hepatic P450 enzymes have also been exploited

CHF2

CH2SH

CH2SSCH2CH3 (prodrug)

CH2C=CH

CH=C=CH2

OCOCH2CH3

CH2SH

Exemestane

CH2

Figure 7.32. Structures of

selection of mechanism-based inactivators of

aromatase.

Inhibition of Cytochrome P450 Enzymes

289

in the design and synthesis of mechanism-based

aromatase inactivators. Substitution of the nor-

mally hydroxylated methyl group (C-19) with a

propargylic or allenic moiety (Figure 7.32) con-

verts the sterol into an irreversible aromatase inhi-

bitor^^^"^^^. The details of aromatase inactivation

by these acetylenic and allenic agents remain

unclear, but it is likely that they are activated

to intermediates that react with either the heme

or the protein (see Sections 3.1 and 3.3.2).

Replacement of the C-19 methyl with a difluo-

romethyl also yields a mechanism-based inactiva-

tor that must be activated by C-19 hydroxylation

(Figure 7.32)^^^^^^ as tritium release from the tri-

tium-labeled difluoromethyl derivative is required

for enzyme inactivation^^^. It is likely that the

difluoromethylalcohol thus produced decomposes

to the acyl fluoride that irreversibly binds to a pro-

tein nucleophile.

The 19-substituted analog of androst-4-ene-3,

17 dione steroid inhibitors, Org-30958 [19-

(ethyldithio)androst-4-ene-3,17-dione], has been

assessed in Phase I clinical trials for estrogen-

dependent breast cancer chemotherapy^^'*. The

ethyldithio substitution apparently renders the

steroid more stable extracellularly than the free

thiol Org-30365 (19-mercapto-androst-4-ene-3,

17-dione), resulting in vivo in animal models in an

8-fold

greater aromatase inhibitory activity than

either 4-OHA or SH-489. Its in vivo potency

requires intracellular reduction of the disulfide to

release the 19-mercapto analog Org-30365, a

more potent mechanism-based human placental

aromatase inactivator"^^^ than 4-OHA or

SH489494.

Clii^ically effective mechanism-based aro-

matase inactivators can also be obtained by intro-

ducing substituents at the 4- or 6-positions of the

sterol skeleton. 4-Acetoxy- and 4-hydroxy-

4-androstene-3,17-dione (4-OHA) (Figure 7.32)

irreversibly inactivate placental aromatase by

catalysis-dependent mechanisms involving the 19-

methyl^^^'

^^'*.

A possible mechanism for inhibition

of aromatase by the 4-substituted analogs, as illus-

trated by 4-OHA, is shown in Figure 7.33. 4-OHA

is used for the treatment of estrogen-dependent

breast cancer^^^' ^^^. Of a series of A^'^, A"^'^,

and A^'^ analogs evaluated as prospective aro-

matase inhibitors in preclinical trials, FC 24928

(4-aminoandrostan-l,4,6-triene-3,l7-dione) is the

most promising candidate because it inactivates

human placental aromatase activity as potently as

4-OHA and FCE-24304 (6-methylene-androstan-

l,4,-diene-3,l7-dione) but, unlike both these

compounds, it has little intrinsic androgenic

activity and does not affect 5a-reductase or

P450 516-518^

sec

Conjugation of the 4-hydroxyandrostene

nucleus as in 1,4,6 androstatriene-3,17-dione

(ATD),

conveys aromatase inhibitory and marked

tumor regression activities (—80%)^^^'

^^^.

On the

other hand, the introduction of a C^-methyl into

l,4-androstadiene-3,l7-dione as in Atamestane

(1 -methylandrosta-1,4-diene-3,17-dione, SH-489),

apparently enhances its affinity

(K^

~2 nM vs

of 29 nM for 4-OHA) for the human placental aro-

matase while slowing its inactivation of the

enzyme, thereby reducing the production of estro-

genic products^ ^^' ^^^. The compound along with

its 1,2 methylene-substituted congeners has been

evaluated in Phase I clinical trials for possible

therapy of estrogen-dependent conditions such

as breast cancer and benign prostatic hypertrophy.

Additional steroidal agents explored for their

aromatase suicide inactivation include androst-

5-ene-7,l 7-dione and its 19-hydroxy derivative^^^.

Turnover-dependent irreversible inactivation of

the enzyme via protein modification is also

achieved by introducing a 6-keto group into the

steroid skeleton (Figure 7.33)^^^"^^^. Monitoring

the ^H:

^^^C

ratio in studies with the C-19 double-

labeled inhibitor indicates that the C-19 methyl,

one of the C-19 hydrogens and, from a separate

double label experiment, the Ip-hydrogen, are

retained in the covalently bound species^^^. These

findings do not define the underlying inactivation

mechanism but appear to exclude the involvement

of C-19 demethylation and aromatization,

although normal aromatization is possible because

6-oxoestrone and 6-oxoestradiol are concurrently

formed.

Exemestane, 6-methylene-androsta-1,4-diene-

3,17-dione (Figure 7.32), is an aromatase inhibitor

with an

IC^Q

for inhibition of human placental

aromatase comparable to that of 4-OHA^^^' ^^^.

The

(nM) and

1^2

(min) values for the inactiva-

tion processes were 26 ± 1.4 and 29.0 ± 7.5, and

13.9 ± 0.7 and 2.1 ± 0.2, for Exemestane and

4-OHA, respectively. In spite of its relatively

slow inactivation of aromatase, Exemestane is a

more potent agent in experimental animals^ ^^, and

also much more effectively causes regression of

290

M.A. CorreJa and P.R. Ortiz de Montellano

[Fe]

HO-V

4-Hydroxyandrostenedione

(4-HOA)

Figure 7.33.

possible mechanism for the inactivation of

aromatase

by 4-OHA.

chemically induced mammary tumors in rat

than MDL18962 or SH-4895i^ Because of its

oral effectiveness and potent irreversible aro-

matase inhibition, Exemestane has been recently

approved and introduced into the global market

under the name Aromasin^^^^^ •.

MDL 18,962 10-(2-propynyl)estr-4-ene-3,17-

dione is one of the most potent of the mechanism-

based inactivators of aromatase, with

values of

the order of 3-4 nM, and ty2 of 9 and 6 min,

respectively, for the human and baboon placental

aromatases^^^'

^^"^^ ^^^^ ^^^' ^^^

pj^ase I clinical trials

and preclinical findings indicate that MDL 18,962

is indeed effective in lowering estrogen levels^^^.

The time-dependent inactivation of aromatase

by 10-hydroperoxy-4-estrene-3,17-dione, reportedly

a mechanism-based inactivator of

aromatase^^'*'

^^^,

is inhibited by NADPH or an alternative substrate

and is partially reversed by dithiothreitol. Other

high affinity 10-substituted analogs, such as the

mechanism-based inhibitor 10-P-mercaptoestr-

4-ene-3,l7-dione, and the competitive inhibitors

(19i^)-10-oxiranyl and lO-thiiranyl-estr-4-ene-

3,17-diones and 10-p,-MeSCH2-estr-4-ene-3,17-

dione, have been reported^^i, 492,494,536,537

Q_IQ^

C-2 hydroxyethyl bridged steroids synthesized as

stable carbon analogs of the 2p-hydroxylated 19-

oxoandrostenedione, a putative intermediate in the

aromatization reaction, have been found to be

potent competitive inhibitors of

the

human placen-

tal aromatase^^^. In contrast, related halohydrin

analogs are potent mechanism-based aromatase

inhibitors^"^^.

Finally, 4p,5p-epoxyandrostenedione as well

as its 19-hydroxy and 19-oxo derivatives have

been examined as aromatase inhibitors. The epox-

ides were found to be weak competitive inhibitors,

whereas the 19-hydroxy and 19-oxo derivatives

were largely ineffective^^^.

5.3. Lanosterol 14-Demethylation

The 14-demethylation of lanosterol is a

key step in the biosynthesis of cholesterol. The

preferential inhibition of the P450 enzyme that

catalyzes this reaction by a number of substituted

imidazoles, pyridines, pyrimidines, and other

lipophilic heterocycles has been exploited in the

construction of clinically important antifungal

agents^'

The antifungal action following the

inhibition of 14-demethylase is thought to result

from the accumulation of 14-methyl sterols in the

membranes of susceptible ftmgi that cause deleteri-

ous changes in their membrane permeability^"^^"^^^.

Miconazole (Figure 7.34), fluconazole

(Figure 7.34), ketoconazole (Figure 7.1), and

Inhibition

Cytochrome

P450

Enzymes

291

COCH3

CH(0H)CH3

/=/

O-

Miconazole

N-N

Fluconazole

Figure

7.34.

Competitive inhibitors

lanosterol

14-demethylase.

The

group

R on the

sterols

either

C(CH3)CH2CH2CH2C(CH3)2

or a

variant.

Fluconazole

and

miconazole

are

clinically

employed

antifungal

agents.

related azole structures inhibit the fungal 14-

demethylase activity at extremely low (nM) con-

centrations^"*^' ^^'^^ but only inhibit the

14-demethylase activity of the mammalian host at

higher (up to 100

|JLM)

concentrations^"^^'

^^^.

Low

doses of ketoconazole, however, appear to inhibit

not only the C^^ 20 ty^se, but also human hepatic

CYP3A4, and thus this agent is not very clinically

useful^^^' ^^^ As expected, the substituted imida-

zoles and other nitrogen-based heterocyclic

antifungal agents coordinate with the P450 heme

iron to yield typical type II binding spectra^^^.

Potent isoprenoid-containing imidazole

antifungal agents, such as AFK-108 (l-[2-

(2,4-dichlorophenyl)-2-((2£)-3,7-dimethylocta-

2,6-dienyloxy) ethyl]-IH-imidazole), have also

been developed whose geranyl moiety specifically

interacts with the sterol side-chain recognition

region within the 14a-demethylase active site.

Unfortunately, this compound and its famesyl

derivative inhibited not only the rat liver counter-

part but also other hepatic drug metabolizing

P450 enzymes, placing their practical utility in

question^^^.

Structural considerations have also prompted the

development of 14a-methyl-15-aza-D-homosterols

(Figure 7.34), which exhibit relatively high in

vitro fungistatic potencies and high fungicidal

activities as well as respectable in vivo efficacies

in a murine model of candidiasis^^'^. Introduction

of an oxime functionality (=NOH) at position 15

of the sterol group, also results in an inhibitor with

a very similar geometry of the nitrogen as that in

the 15-azahomosterols^^^. Similarly, 7-oxo-24,25-

dihydrolanosterols, which uncouple the catalytic

turnover of CYP51 by competitively blocking

transfer of a second electron to the oxyfer-

rocomplex, have also been considered as prospec-

tive inhibitors^^^. Furthermore, 15-, 32-, and

15,21-oxylanosterol analogs have been examined

as mechanism-based inactivators of lanosterol

demethylase. Of these (32*S)-vinyllanost-8-en-

3p,32-diol is a potent time-dependent inactivator

(^inac/^i = 0.36 min-^

|JLM-^),

while the (32i?)-

vinyllanost-8-en-3p,32-diol functions solely as a

competitive demethylase inhibitor^^^. Replace-

ment of the 15-hydrogen normally lost during

catalysis by a fluorine yields 15a-fluorolanost-

7-en-3p-ol which is oxidized by the enzyme to the

14-aldehyde but proceeds no further due to the

fluorine substitution^^^. The compound is there-

fore a "prodrug" inhibitor that requires metabolic

activation of its latent form rather than a true

mechanism-based inactivating agent.

Steroid analogs with amino and imidazole

functions attached to the 14-methyl carbon have

292 M.A. Correia and P.R. Ortiz de Montellano

been developed as inhibitors of lanosterol 14a-

demethylation (Figure 7.34)^^^. Steroids with a

14-(l-hydroxyethyl), 14-(l-oxoethyl), or 14-car-

boxyl group are competitive inhibitors of the

enzyme (Figure 7.34)^^^"^^^. Lanosterol analogs

in which the 14a-methyl group has been replaced

by a vinyl, ethynyl, allyl, propargyl,

1-hydroxy-

propargyl,

1-ketopropargyl,

difluoromethyl,

epoxide, or episulfide moiety have been synthe-

sized as potential mechanism-based inhibitors of

lanosterol 14a-demethylase^^^~^^^. Although the

ethynyl sterols are clearly mechanism-based inac-

tivators of the enzyme^^^, most of these com-

pounds act as simple competitive inhibitors (e.g.,

the epoxide and episulfide analogs). Finally,

because of their cholesterol lowering potential,

14a-demethylase inhibitors have also been

considered as hypolipidemic agents^^^.

5.4. Other Biosynthetic

Sterol Hydroxylases

A current strategy for the development of new

drugs for the treatment of prostatic cancer is the

development of inhibitors of CYP17-dependent

androgen biosynthesis. Thus, as in the case of

other steroidogenic enzymes, an intense search

has been undertaken for reversible and irreversible

CYP17 inhibitors^^^. The resulting agents include

substituted imidazoles, pyridines, pyrimidines,

and other lipophilic heterocycles as the enabling

moiety^

Of these, abiraterone

(17-(3-

pyridyl)androsta-5,16-dien-3p-ol), a potent inhi-

bitor (IC5Q, 4 nM) of the hydroxylase activity of

human cytochrome CYP17, has been employed

clinically^^^ Its potency is apparently due to acti-

vation of the 16,17-double bond that is required

for irreversible binding of these pyridyl steroids to

CYPn^^^ Analogs such as [17-(5-

pyrimidyl)androsta-5,16-diene-3p-ol] and its 3-

acetyl derivative, which are even more potent

CYP17 inhibitors in rats than abiraterone, are par-

ticularly promising^^^. With testicular microsomal

CYP17,

these compounds apparently exhibit the

dual characteristics of a type II binding spectrum

and noncompetitive inhibition^^^

Another mechanism-based inactivator of

both the 17-hydroxylase and C-17/C-20 lyase

activities of CYP17 is 17p-(cyclopropylamino)-

androst-5-en-3p-ol^^^. This compound reportedly

does not inhibit P450g^^ or the sterol 21-hydroxy-

lase.

In contrast, sterols with a 17-difluoromethyl

group selectively inactivate the C-21 hydroxylase,

whereas sterols with a 17-dichloromethyl, -vinyl,

or -ethynyl function inactivate both the CI7- and

C21-hydroxylases^^. More recently, 20-fluoro-

17(20)-pregnenolone derivatives were designed as

enol mimics of pregnenolone. All of the targeted,

novel fluoroolefins were found to be potent

inhibitors of C-17(20) lyase579.

Similar strategies have been used to develop

inhibitors that target the C-18 hydroxylation

involved in the biosynthesis of aldosterone^^^"^^^.

Thus,

aldosterone analogs with C-18 iodomethyl,

chloromethyl, allyl, propargyl, vinyl, and methyl-

thiomethyl functionalities have been synthesized

and some have been found to irreversibly inactivate

the enzyme. An active site-directed 18-acetylenic

deoxycorticosterone [21 -hydroxy-13 (-2-)propy-

nyl)-18-nor-preg-4-ene-3,20-dione, MDL19,347]

is a promising inactivator of the rat and rhesus

monkey adrenal corticosterone 18-hydroxylase

(K^ ^ 38 nM; f,/2, 4.6 min) that also effectively

reduces plasma aldosterone levels in

vivo^^^.

This

compound appears to be a relatively selective

inhibitor of aldosterone biosynthesis, as it fails to

inhibit the

p-hydroxylation of corticosterone and

j)Q^586 Rationally developed inhibitors such

as these may be useful in the management of con-

ditions such as hypertension, hypokalemia, and

edema that are associated with hyperaldosteronism.

5.5. Fatty Acid and Leukotriene

Monooxygenases

Microsomal P450 enzymes of the CYP2C and

4A subfamilies in tissues such as the liver, lung,

kidney, peripheral vasculature, and polymor-

phonuclear leukocyte oxidize a variety of fatty

acids,

including arachidonic acid and its deriva-

tives,

to physiologically active metabolites.

Some of these metabolites are well recognized

as biologically important regulators of renal,

pulmonary, and cardiac function and vascular

tone^^^"^^^.

Two general classes of such vaso-

active metabolites are known to be produced in

vascular and extravascular tissues: EETs gener-

ated by CYP2C epoxygenases, and HETEs

products hydroxylated at the o)- and w-l positions

Inhibition of Cytochrome P450 Enzymes

293

by CYP4A^^^-^^^. Furthermore, with the excep-

tion of 20-HETE, all the other arachidonic acid

products occur as stereo- and regioisomers that

vary considerably in their biological activities and

potencies^92, 593

jj^^s^

55.^ 8,9-, 11,12-, and

14,15-EETs are potent vasodilators, specially in

various capillary beds, that act through activation

of K+-channels in vascular smooth muscle cells.

In contrast, 12(i^)- and 20-HETEs are potent vaso-

constrictors. 20-HETE is a particularly potent

cerebral and renal microvessel vasoconstrictor, as

well as a mediator of other important physiologi-

cal processes (reviewed in Chapter 11). The

altered production of EETs and 20-HETE in vari-

ous genetic and experimental models of patholog-

ical diseases has led to their consideration as

plausible causative factors. Thus, not surprisingly,

the P450 enzymes responsible for their biosynthe-

sis have been singled out as key targets not only

for defining the pathological roles of these

metabolites, but also as possible sites for pharma-

cological intervention in the treatment of these

conditions^^^"^^^. In this context, both the previ-

ously existing and new agents have been exploited

as reversible or irreversible inhibitors of these

enzymes.

Accordingly, two reversible inhibitors of

CYP4A enzymes have been assessed as probes of

20-HETE involvement in various physiological

processes with varying degrees of success: the not

so selective antifungal agent miconazole (Section

2.3)^^^"^^^

and the rationally designed and thus

more selective, A/-methylsulfonyl-12,12-dibro-

mododec-11-enamide (DDMS; Figure 7.35) and

its acid analog, 12,12-dibromo-dodec-ll-enoic

(DBDD)602-^oi DDMS has been found to be an

effective acute and chronic blocker of 20-HETE

formation in vivo in rats^^^"^^"^. More recently an

even more potent and selective inhibitor of 20-

HETE formation, HET0016 (A^-hydroxy-A^'-(4-

butyl-2-methylphenyl)formamidine) (Figure 7.35)

has been discovered through combinatorial chem-

istry, with an

IC3Q

of 35 ±4

nM^^^.

Because of its

high selectivity, it holds considerable promise as

a probe of 20-HETE dependent function.

Several mechanism-based irreversible inhi-

bitors of P450-dependent arachidonic acid metab-

olism have also been developed. For instance, in

11-dodecynoic acid, the terminal acetylenic analog

of lauric acid, and 10-undecynoic acid have been

shown to inactivate hepatic CYP4A lauric acid

(^-hydroxylases while minimally altering the

spectrophotometrically detectable hepatic P450

content or function^^'

^°^'

^^^.

The high susceptibility of these acetylenic

fatty acid inactivators to p-oxidation, however,

compromises their in vivo utility and prompted the

development of 10-undec5niyl sulfate (10-SUYS),

the acetylenic analog of 10-undecynoic acid in

which the carboxyl group is replaced by a sulfate

(Figure 7.35)^^^. The sulfate is not susceptible to

P-oxidation and thus prolongs the biological

lifetime, reduces the toxicity, and enables the

use of sulfate agent as an in vivo mechanism-based

CONHSO2CH3

DDMS HET0016

x-^ .OH

N N

.CO2H

17-ODYA

.OSOgH

10-SUYS

.CONHSO2CH3

MSPPOH

Figure 7.35. Reversible and mechanism-based inhibitors of enzymes involved in the metabolism of fatty acids and

related endogenous substrates.

294

M.A.

Correia

and P.R.

Ortiz

Montellano

CYP4A inactivator^^^. Administration of a

single dose of 10-SUYS, a potent and selective

mechanism-based CYP4A inactivator, acutely

reduced the mean arterial blood pressure as well

as the urinary 20-HETE excretion in sponta-

neously hypertensive rats, consistent with the

inactivation of renal 20-HETE formation^ ^^

These findings thus suggest that 20-HETE could

play an important role in blood pressure regulation

in hypertensive states and that the inhibition of

its synthesis in these conditions may be of thera-

peutic benefit^ ^^

CYP4A and CYP2C isoforms are also

inactivated by the acetylenic fatty acid analog,

17-octadecynoic acid (17-ODYA; Figure 7.35),

in a mechanism-based process^ ^^~^^'*. 17-ODYA

has been particularly useful in vivo to probe the

involvement of these metabolites in physiological

and/or pathological processes^^v, 598, 603, 605-609,

612-618 ji^g ygg jj^g helped elucidate the specific

role of 20-HETE in the myogenic activation and

h)^oxic dilation of skeletal muscle resistance

arteries as well as the vasodilatory effects of NO

in the renal microcirculation^^^'

^^^' ^^'^'

^^l

The acetylenic fatty acids 15-hexadecynoic

(15-HDYA) and 17-ODYA have also been

explored as modulators of leukotriene B^ (LTB^),

an important and clinically relevant inflamma-

tory mediator, and its physiologically active

(o-hydroxylated metabolite^i^'^22. Both 15-HDYA

and 17-ODYA inactivated the polymorphonuclear

leukocytic LTB^ w-hydroxylase in whole cells

and cell lysates^^^. In contrast, the shorter-chain

acid, 10-undecynoic acid was much less effective,

while the saturated analogs of 15-HDYA and

17-ODYA were inactive. 15-HDYA and 17-

ODYA also inactivate pulmonary prostaglandin

(o-hydroxylases^^^.

The mechanism-based inactivator 1-ABT

(Section 3.3.5), which is not very selective and

inactivates multiple P450 enzymes, including

those responsible for the synthesis of both EETs

and 20-HETE, has also been used, alone or in con-

junction with other inhibitors, to assess the role of

these metabolites in skeletal muscle angiogenesis

induced by electrical stimulation as well as in

the renal and vasoconstrictor actions of angio-

tensin

IP'^^'

602,

608,

624^

Mechanism-based inhibitors of EET formation

have also been used as probes of the physiological

roles of these metabolites. One such agent is

iV-methylsulfonyl-6-(2-propargyloxyphenyl)hexa-

namide (MSPPOH; Figure 7.35)^^^'

^25-627

Finally, since fatty acid hydroxylations at the

0),

(0-1, (0-2, (0-3, and (o-4 positions also occur in

plants and bacteria^^^'

^^^,

acetylenic fatty acids

have also been employed to examine whether

these hydroxylases are mechanistically similar

to their mammalian counterparts and/or to iden-

tity the role of specific plant and bacterial P450

enzymes^^^' ^^^. Thus, midchain and terminal

acetylenes such as 10-dodecynoic acid have been

used as probes of plant lauric acid (o—hydroxy-

lases^^'6^^.

Similarly, 17-ODYA has been shown to

inactivate

P450BJ^.3

(CYP108), an enzyme that

hydroxylates fatty acids at positions other than the

terminal ((o) carbon, through a heme alkylation

mechanism^^o.

6. Summary

Our knowledge of the structure and function

of P450 enzymes has greatly expanded over the

past three decades, as has our appreciation of

the chemical functionalities that they accept as

substrates and of the moieties that can inhibit or

terminate their catalytic action. As a result, multi-

ple routes are now available for the inhibition or

inactivation of P450 enzymes that target either the

protein and/or the heme. This knowledge has been

exploited in the design and construction of both

reversible and irreversible inhibitors of increasing

potency, specificity, and potential practical utility.

One area of continuing growth is that of reversible

inhibitors that, by simultaneously conforming to

the predominantly lipophilic contours of the P450

active site and providing a nitrogen that coordi-

nates to the heme iron atom, have achieved

enhanced potency and specificity. These agents

include the clinically relevant azole antifungal and

cancer chemotherapeutic drugs (Section 5).

The high promise of irreversible P450

inactivating agents has not yet culminated in

agents of large-scale practical utility. Progress

in the field has continued to provide instruction on

approaches for the design of mechanism-based

inhibitors that inactivate a P450 enzyme with high

specificity and negligible release of reactive

metabolites into the medium, whether the inacti-

vation involves modification of the protein or the